论文部分内容阅读
为使具有诸多优点的互补金属氧化物半导体(CMOS)传感器更适合空间高分辨成像,寻求空间高分成像的新型技术,提出了更利于微光成像和推扫成像的卷帘数字域时间延迟积分(TDI)算法。同时研究数字域TDICMOS相机成像质量,详细分析了其噪声来源和特性,并建立了数字域积分图像信噪比(SNR)与积分级数的关系模型,讨论了积分时间和光照度对SNR的影响。最后利用设计的IBIS5-B-1300卷帘数字域TDI CMOS原理样机开展验证实验。实验结果表明本文算法能明显提高成像质量,数字域10级积分图像SNR由未积分的19.07dB提高至29.21dB,而且级数越大,SNR越大。理论分析和实验验证均表明M级卷帘数字域TDI可使图像SNR提高M(σAD+σCMOS)/(MσAD+槡MσCMOS)倍,其中σAD和σCMOS与选择的CMOS传感器有关,另外σCMOS还受积分时间和光照度的影响。
In order to make the complementary metal-oxide-semiconductor (CMOS) sensor with many advantages more suitable for space high-resolution imaging and seek for a new technology of high-spatial-spatial imaging, a rolling shutter digital domain time-delay integral which is more favorable to low-light imaging and push- (TDI) algorithm. At the same time, the imaging quality of digital TDICMOS camera was studied. The sources and characteristics of TDICMOS digital camera were analyzed. The relationship model between signal-to-noise ratio (SNR) and integral series was established. The influence of integral time and illuminance on SNR was also discussed. Finally, the design of IBIS5-B-1300 shutter digital domain TDI CMOS prototype prototype validation experiments. Experimental results show that the proposed algorithm can significantly improve the imaging quality. The SNR of the 10-level integral image in the digital domain increases from 19.07dB, which is not integrated, to 29.21dB. The larger the number of stages, the larger the SNR. Both theoretical analysis and experimental verification show that the digital SNR of M-roll blind digital domain can increase the SNR of the image by M (σAD + σCMOS) / (MσAD + 槡 MσCMOS) times, where σAD and σCMOS are related to the selected CMOS sensor and σCMOS is also affected by the integration time And the impact of light intensity.